13 - The forecast that lies

There is a moment in every enterprise forecast review that most leaders recognise.

The spreadsheet is open. The pipeline looks full. The rep is talking through each deal with confidence. And somewhere in the room, a quiet thought forms: this number is not real.

Nobody says it out loud. The rep has relationships. The deals are large. Challenging the forecast feels like challenging the person. So the meeting ends. The number goes up to the board. And three months later, the quarter comes in short.

This is not a talent problem. It is a structural one.

The standard most teams are using is wrong

When something goes wrong with a forecast, most managers ask the same question: why did the rep not flag this earlier?

Wrong question.

The rep flagged it the only way the system allowed. They put a close date in the CRM and called it committed. There was no other mechanism. No entry condition. No evidence standard. No distinction between a deal that is progressing and a deal someone believes in.

In enterprise sales, a deal can sit in late stage for six months and still be real. Legal takes time. Budget cycles are fixed. Executive alignment moves slowly. None of that makes the deal fictitious.

But it does mean weekly movement is the wrong signal to track.

The right standard is this: what is the last thing the prospect did that cost them something?

Time spent in an evaluation. A technical team assigned internally. An executive sponsor who got on a call to reconfirm priority. Legal who actually opened the contract. These are costly actions. They require the buyer to spend internal capital. A prospect who is quietly losing interest does not spend capital like this.

If you cannot name a costly buyer action, you do not have a forecast. You have a hope with a close date on it.

I use this question in every engagement now. It takes about ten minutes to expose six months of wishful thinking.

Why the problem does not fix itself

Most companies treat this as a rep problem. Reps inflate. Reps are optimistic. Reps avoid hard conversations.

All true. But it is the wrong diagnosis.

The real issue is that every layer in the chain has a structural incentive to make the number look better than it is.

The rep does not want to show a thin pipeline. The manager does not want a difficult conversation with someone who has been nursing a deal for nine months. The VP does not want to deliver a short number to the CRO. The CRO does not want to explain it to the board. So the number moves upward, quietly inflated at each handoff, and nobody formally owns the distortion.

I worked inside a business doing over 80 million in revenue. Good team, experienced leaders. When I asked which deals in the committed forecast had confirmed executive sponsorship, nobody could answer with specificity. Not because they were hiding it. Because the system had never required them to know.

The deals had names. They had values. They had close dates. What they did not have was a verified condition of reality.

That is the actual problem. Not the reps. The structure that made optimism the path of least resistance.

Three structural fixes

1. Write the entry condition for every stage

Every pipeline stage should have one written condition. Not a label. A specific buyer action that must have occurred before a deal can move there.

Stage four does not mean proposal submitted. It means: proposal submitted, commercial conversation held, and a decision-maker has confirmed a review timeline exists. If those conditions are not true, the deal does not belong there regardless of how long the rep has been working it.

I worked with a company that had nine stages and zero definitions. Every rep interpreted them differently. Their forecast was not a number. It was an average of nine different opinions.

We wrote definitions in a single afternoon. Within one quarter, forecast accuracy improved from 51 percent to 74 percent. The pipeline did not get bigger. It got honest.

2. Separate pipeline from committed forecast

Pipeline is everything in play. Forecast is what a rep is prepared to stake their credibility on closing within the period.

These are not the same number. Most companies collapse them into one figure and wonder why visibility is poor.

Create a commit column. At each review, ask every rep to separate the two. A deal in pipeline is being worked. A deal in commit is one they will own if it does not close.

When I install this, the first conversation is always uncomfortable. Reps are not used to being asked to commit to a number without a safety net. That discomfort is information. A rep who cannot commit to anything is telling you the pipeline is thin. You need to know that in week two, not week eleven.

One company I worked with had 4.2 million in pipeline and committed to 1.1 million when we ran this exercise. Leadership thought they had a 3 million quarter. They had 900k. Better to find out in planning than at close.

3. Change the question in every deal review

Most enterprise forecast reviews are structured as status updates. The rep talks. The manager listens. Activity gets confirmed. Almost nothing gets challenged.

Change the question. For every deal in commit, ask: what has the buyer done on their side that confirms this is still a priority?

Not what the rep has sent. Not what the rep has planned. What the buyer has done.

A buyer who has assigned internal resources, moved their own people onto the evaluation, or escalated the project internally is a buyer who is buying. A buyer who is polite and responsive but has taken no internal action is a risk, whatever the rep believes.

This question is uncomfortable the first time you ask it in a review. By the third time, the team starts preparing for it. They go back to their prospects and create the conditions that generate a real answer. That behavioural shift alone changes forecast quality faster than any tool or training programme.

The forecast does not lie on its own

The structure lets it.

When there is no evidence standard, optimism fills the gap. When pipeline and forecast are the same number, there is no accountability layer. When the review question is about activity rather than buyer behaviour, nothing gets exposed until it is too late.

Fix the structure. The number will follow.